Learning from Time-Changing Data with Adaptive Windowing

TLDR: A new approach for dealing with distribution change and concept drift when learning from data sequences that may vary with time is presented, using sliding windows whose size is recomputed online according to the rate of change observed from the data in the window itself.

Abstract: We present a new approach for dealing with distribution change and concept drift when learning from data sequences that may vary with time. We use sliding windows whose size, instead of being fixed a priori, is recomputed online according to the rate of change observed from the data in the window itself: The window will grow automatically when the data is stationary, for greater accuracy, and will shrink automatically when change is taking place, to discard stale data. This delivers the user or programmer from having to guess a time-scale for change. Contrary to many related works, we provide rigorous guarantees of performance, as bounds on the rates of false positives and false negatives. In fact, for some change structures, we can formally show that the algorithm automatically adjusts the window to a statistically optimal length. Using ideas from data stream algorithmics, we develop a time- and memory-ecient version of this algorithm, called ADWIN2. We show how to incorporate this strategy easily into